Calendarize

A schedule-to-calendar converter that extracts structured events from unstructured sources — screenshots, PDFs, images, emails or pasted text — and exports them as standard calendar files.

Multi-format ingestion via drag-and-drop upload or direct text paste; AI-driven extraction of dates, times, titles and locations from documents with no fixed layout; a mandatory review stage where every proposed event can be edited, adjusted or removed before anything is committed; .ics export compatible with Google Calendar, Apple Calendar and Outlook; optional accounts that persist conversion history; guest use with no signup required.

A Next.js application pairing an LLM extraction pipeline (OpenAI) with Zod schema validation, so model output is parsed into a typed event structure and rejected when it does not conform — keeping malformed AI responses out of the database. Prisma manages persistence for user accounts and saved conversion history; Tailwind and Framer Motion handle the review interface, where clarity matters more than density because users are verifying machine output against a source document.

The product's defining constraint is that extraction results are never written directly to a calendar. The AI proposes; the user confirms. This is a deliberate design decision rather than a limitation — an incorrect event silently added to a real calendar is worse than no automation at all, because it removes the user's reason to double-check.